Would it be worth making a dll for this?
 

Developed an Excel add-in that uses a number of heavy (on memory and
processor) array manipulations.
Would there be any advantage in making a dll to let this do the work? It
would be given an array
and a number of other arguments and the return will usually be an array. I
suppose the only way of knowing will be
to try it, but I am interested in some general thoughts about this.

RBS

Would it be worth making a dll for this?
 

A dll will be more efficient as it is compiled code, against an add-in's
interpretive. If the calculations are as heavy as you say, it probably will
help. But offset against that is distribution issues of course, not
insurmountable, but greater than a simple add-in.

--

HTH

RP
(remove nothere from the email address if mailing direct)


"RB Smissaert" wrote in message
...
Developed an Excel add-in that uses a number of heavy (on memory and
processor) array manipulations.
Would there be any advantage in making a dll to let this do the work? It
would be given an array
and a number of other arguments and the return will usually be an array. I
suppose the only way of knowing will be
to try it, but I am interested in some general thoughts about this.

RBS

Would it be worth making a dll for this?
 

The manipulations are usually just loops, comparing values (is element a
same as element b,
or is element a element b) and writing values to a new array.
These arrays can be big though, say 100.000 rows and maybe 4 columns.
The other advantage there could be is in case I would make this add-in
commercial.
As you can't see the code of the .dll file there is no problem with
protection as there is with .xla files.
I am not really sure how this exactly would work, but I understand you could
make a key that would be
needed to unlock the .dll file. Would be interested in any thoughts about
this.

RBS

"Bob Phillips" wrote in message
...
A dll will be more efficient as it is compiled code, against an add-in's
interpretive. If the calculations are as heavy as you say, it probably
will
help. But offset against that is distribution issues of course, not
insurmountable, but greater than a simple add-in.

--

HTH

RP
(remove nothere from the email address if mailing direct)


"RB Smissaert" wrote in message
...
Developed an Excel add-in that uses a number of heavy (on memory and
processor) array manipulations.
Would there be any advantage in making a dll to let this do the work? It
would be given an array
and a number of other arguments and the return will usually be an array.
I
suppose the only way of knowing will be
to try it, but I am interested in some general thoughts about this.

RBS

Would it be worth making a dll for this?
 

Dave Scarmozzino has a nice product that allows time-limited trials, license
keys etc. It is called AppSentinel and can be found at
http://www.thescarms.com/appsentinel/default.asp

--

HTH

RP
(remove nothere from the email address if mailing direct)


"RB Smissaert" wrote in message
...
The manipulations are usually just loops, comparing values (is element a
same as element b,
or is element a element b) and writing values to a new array.
These arrays can be big though, say 100.000 rows and maybe 4 columns.
The other advantage there could be is in case I would make this add-in
commercial.
As you can't see the code of the .dll file there is no problem with
protection as there is with .xla files.
I am not really sure how this exactly would work, but I understand you
could
make a key that would be
needed to unlock the .dll file. Would be interested in any thoughts about
this.

RBS

"Bob Phillips" wrote in message
...
A dll will be more efficient as it is compiled code, against an add-in's
interpretive. If the calculations are as heavy as you say, it probably
will
help. But offset against that is distribution issues of course, not
insurmountable, but greater than a simple add-in.

--

HTH

RP
(remove nothere from the email address if mailing direct)


"RB Smissaert" wrote in message
...
Developed an Excel add-in that uses a number of heavy (on memory and
processor) array manipulations.
Would there be any advantage in making a dll to let this do the work?
It
would be given an array
and a number of other arguments and the return will usually be an
array.
I
suppose the only way of knowing will be
to try it, but I am interested in some general thoughts about this.

RBS

Would it be worth making a dll for this?
 

Thanks for tip, will keep it in mind.

RBS


"Bob Phillips" wrote in message
...
Dave Scarmozzino has a nice product that allows time-limited trials,
license
keys etc. It is called AppSentinel and can be found at
http://www.thescarms.com/appsentinel/default.asp

--

HTH

RP
(remove nothere from the email address if mailing direct)


"RB Smissaert" wrote in message
...
The manipulations are usually just loops, comparing values (is element a
same as element b,
or is element a element b) and writing values to a new array.
These arrays can be big though, say 100.000 rows and maybe 4 columns.
The other advantage there could be is in case I would make this add-in
commercial.
As you can't see the code of the .dll file there is no problem with
protection as there is with .xla files.
I am not really sure how this exactly would work, but I understand you
could
make a key that would be
needed to unlock the .dll file. Would be interested in any thoughts about
this.

RBS

"Bob Phillips" wrote in message
...
A dll will be more efficient as it is compiled code, against an add-in's
interpretive. If the calculations are as heavy as you say, it probably
will
help. But offset against that is distribution issues of course, not
insurmountable, but greater than a simple add-in.

--

HTH

RP
(remove nothere from the email address if mailing direct)


"RB Smissaert" wrote in message
...
Developed an Excel add-in that uses a number of heavy (on memory and
processor) array manipulations.
Would there be any advantage in making a dll to let this do the work?
It
would be given an array
and a number of other arguments and the return will usually be an
array.
I
suppose the only way of knowing will be
to try it, but I am interested in some general thoughts about this.

RBS